Difference between revisions of "GTC SC11"
(→Experiment simulation parameters) |
(→Performance Results) |
||
Line 78: | Line 78: | ||
This show a trace of a single execution on one MPI process (Multiple nodes/gpus can be utilized as well, performance behavior is similar across each process). | This show a trace of a single execution on one MPI process (Multiple nodes/gpus can be utilized as well, performance behavior is similar across each process). | ||
+ | |||
+ | |||
+ | [[Image:gtc_gpu_profile.png|750px]] | ||
+ | |||
+ | This shows a representative a profile for the GPU kernel. |
Revision as of 01:47, 29 November 2011
Background
In the fall of 2011, several performance studies were conducted on a port of the GTC application of GPUs (full paper).
Experiment setup
[ http://keeneland.gatech.edu/ Keeneland ] was chossen at the site for running these experiments, and is accessible to the developers as well.
The GNU compilers were chosen so as to minimize in conflict with CUDArt with does not support other compilers.
Here is the Makefile used:
# Define the following to 1 to enable build BENCH_GTC_MPI = 1 BENCH_CHARGEI_PTHREADS = 0 BENCH_PUSHI_PTHREADS = 0 BENCH_SERIAL = 0 SDK_HOME = /nics/c/home/biersdor/NVIDIA_GPU_Computing_SDK/ CUDA_HOME = /sw/keeneland/cuda/4.0/linux_binary NVCC_HOME = $(CUDA_HOME) TAU_MAKEFILE=/nics/c/home/biersdor/tau2/x86_64/lib/Makefile.tau-cupti-mpi-pdt-openmp-opari TAU_OPTIONS='-optPdtCOpts=-DPDT_PARSE -optVerbose -optShared -optTauSelectFile=select.tau' TAU_FLAGS=-tau_makefile=$(TAU_MAKEFILE) -tau_options=$(TAU_OPTIONS) CC = tau_cc.sh $(TAU_FLAGS) MPICC = tau_cc.sh $(TAU_FLAGS) NVCC = nvcc NVCC_FLAGS = -gencode=arch=compute_20,code=\"sm_20,compute_20\" -gencode=arch=compute_20,code=\"sm_20,compute_20\" -m64 --compiler-options '-finstrument-functions -fno-strict-aliasing' -I$(NVCC_HOME)/include -I. -DUNIX -O3 -DGPU_ACCEL=1 -I./ -I$(SDK_HOME)/C/common/inc -I$(SDK_HOME)/shared/inc NVCC_LINK_FLAGS = -fPIC -m64 -L$(NVCC_HOME)/lib64 -L$(SDK_HOME)/shared/lib -L$(SDK_HOME)/C/lib -L$(SDK_HOME)/C/common/lib/linux -lcudart -lstdc++ CFLAGS = -DUSE_MPI=1 -DGPU_ACCEL=1 CFLAGSOMP = -fopenmp COPTFLAGS = -std=c99 #CFLAGSOMP = -mp=bind #COPTFLAGS = -fast CDEPFLAGS = -MD CLDFLAGS = -limf $(NVCC_LINK_FLAGS) MPIDIR = CFLAGS += -I$(CUDA_HOME)/include/ EXEEXT = _keeneland_opt_gnu_tau_pdt AR = ar ARCRFLAGS = cr RANLIB = ranlib
PDT was chosen to allow for event filtering here is the select file used:
BEGIN_EXCLUDE_LIST double RngStream_RandU01(RngStream) double U01(RngStream) END_EXCLUDE_LIST
Experiment simulation parameters
Along with the source code 3 sets of simulation parameters were given: A, B, C (largest). Also a choice of m-cell size: 20 or 96 (96 requires significantly more memory). A, B with m-cell size 20 were used for these performance results.
Performance Results
Here are some performance results that show the overall execution model:
This show a trace of a single execution on one MPI process (Multiple nodes/gpus can be utilized as well, performance behavior is similar across each process).
This shows a representative a profile for the GPU kernel.